UNCERTAINTY ASSESSMENT OF THE BLOOD DAMAGE IN A FDA BLOOD PUMP

Chen Song, Vincent Heuveline
2019 Proceedings of the 3rd International Conference on Uncertainty Quantification in Computational Sciences and Engineering (UNCECOMP 2019)   unpublished
Heart failure (HF) is a severe cardiovascular disease, millions people are suffered from HF worldwide. The ventricular assist device (VAD) can replace the function of failing hearts, when there are no heart donations available, it is becoming a common daily practice for the heart failure patients. U.S. Food and Drug Administration (FDA) Critical Path Initiative (CPI) announced a benchmark study of a centrifugal blood pump few years ago in order to improve current practices of applying
more » ... nal fluid dynamics (CFD) to medical devices. In our previous works, we developed our numerical model for the blood pump simulation with the consideration of Uncertainty Quantification (UQ). We introduced a shear layer update approach in order to facilitate and accelerate the moving mesh process in the framework of High Performance Computing (HPC). The uncertainties in the parametric data and geometric information are quantified with the Polynomial Chaos (PC) method, a Multilevel preconditioning technique is therefore proposed for expediting the linear solvers. In this work, we show an instationary blood flow through a FDA blood pump configuration with Galerkin Projection method, which is realized in our open source Finite Element library HiFlow3. We consider the stress-based hemolysis model to demonstrate the blood damage during the operation of the blood pump. Three uncertainty sources are considered: inflow boundary condition, rotor angular speed and dynamic viscosity, the numerical results are demonstrated with more than 45 Million degree of freedoms by using supercomputer.
doi:10.7712/120219.6366.18426 fatcat:ojzer2j7evdstpryv2pcrbfofe